Quick viewing(Text Mode)

Classification of Routinely Processed Anaplastic Large Cell Tumours with a Small Panel of Antibodies

Classification of Routinely Processed Anaplastic Large Cell Tumours with a Small Panel of Antibodies

Histol Histopath (1987) 2: 107-118 and Histopathology

Classification of routinely processed anaplastic large cell tumours with a small panel of antibodies. An immunohistochemical study with clinical follow-up Sonja C. Henzen-Logmansl, Hendrik Mullinkl, Clauss Vennegoorz, Jo Hilgerss, Jan Oortl and Chris J.L.M. Meiierl 'Department of Pathology, Free University Hospital, Amsterdam, The Netherlands; Department of Tumour Immunology, The Netherlands Institute, Amsterdam, The Netherlands; 3Department of Tumour Biology, The Netherlands Cancer Institute, Amsterdam, The Netherlands

Summary. A proportion of anaplastic large cell tumours Introduction is difficult to classify on sections of routinely processed, paraffin-embedded tissue. Differentiation into large cell A number of publications have appeared about the , carcinoma, melanoma or is important diagnostic value of irnmunohistochemistry with monoclonal in order to assess prognosis and proper treatment. Although antibodies in the classification of anaplastic large cell the use of immunohistochemistry has been reported in the tumours (Warnke et al., 1983; Battifora et al., 1984; differentiation between some of these types of , Baumal et al., 1984; Borowitz et al., 1984; Gatter et al., no antibody panel, which can directly differentiate all of 1984, 1985; Lauder et al., 1984; Bosq et al., 1985). In this them, has been described. In the present study we evaluated way e.g. lymphoma could be distinguished from carcinoma the value of a panel of 5 antibodies for the classification or sarcoma (Borowitz et al., 1984), carcinoma from of 29 anaplastic large cell tumours, which could not be lymphoma (Gatter et al., 1984; Lauder et al., 1984; Bosq classified by experienced pathologists using conventional et al., 1985) and melanoma from carcinoma or lymphoma histological and histochemical techniques. The panel, which (Gatter et al., 1985). All these studies have in common that can be used on routinely fixed paraffin-embedded tissue, they were retrospective, did not cover the whole range of consisted of 5 different antibodies directed against keratin, possibilities for these tumours (carcinoma, sarcoma, vimentin, the human milk-fat globule membrane antigen melanoma or lymphoma), and that the results were only MAM-6, a melanoma associated antigen and common rarely evaluated by clinical follow-up. Moreover, in most leucoyte antigen. cases the antibodies used had to be applied on frozen The use of this panel directly resulted in a definite sections. diagnosis in 95% of the cases and provided valuable In the last years we have tested a large number of information for the diagnosis in the remaining cases. The (predominantly monoclonal) antibodies for their relevance diagnosis was confirmed by additional marker studies and in tumour diagnosis. This has resulted in the composition electron microscopy. Moreover, clinical follow-up, of a panel of 5 antibodies, of which 3 are monoclonal, for including treatment data, was in accordance with the the classification of anaplastic large cell tumours and diagnosis based on the panel. applicable on routinely fixed and processed tissue. The panel is composed in such a way that the obtained positive and negative reaction patterns reinforce each other to Key words: Immunohistochemistry - Anaplastic tumours achieve a definite diagnosis. It consists of antibodies against - Large cell tumours - Differential diagnosis keratin (Nagle et al., 1982; Ramaekers et al., 1983; van Muyen et al., 1984), vimentin (Denk et al., 1983; Gabbiani et al., 1983; Osborn et al., 1983; Rungger-Brandle et al., 1983), common-leucocyte antigen (CLA) (Warnke et al., 1983; Lauder et al., 1984; Meijer et al., 1985), a melanoma associated antigen (Duinen et al., 1984; Mackie et al., 1984; Vennegoor et al., 1985), and the human milk-fat globule Offprint request to: S.C. Henzen-Logmans, Department of membrane antigen (HMFGM) MAM-6 (Hilkens et al., Pathology, Free University Hospital, de boelelaan 11 17, 1081 HV 1984). Amsterdam, The Netherlands In this paper we describe, in a prospective study, 7 FEB. 198% 108 lmmunohistochemistry of anaplas tic large cell tumours our results with this panel of antibodies on 29 anaplastic section was included. Phosphate buffered saline (PBS), large cell tumours, which could not be classified by non-immune ascites fluid or nomal rabbit immunoglobulin exprerienced surgical pathologists with conventional were used as negative controls. histology. After establishing the (initial) diagnosis by means of the pane, additional immunohistochemical marker Antibodies studies and electron microscopy were performed to verify it. Moreover, the diagnostic value was evaluated by means The antibodies used in the panel and their specificities of clinical follow-up including treatment results. It will be are described in Table 2. shown that the use of our antibody panel resulted in a The antibodies that were subsequently used to verify definite diagnosis in the vast majority of the tumours while the diagnosis are listed in Table 3. in only one case additional marker studies were needed. Goat anti-mouse Ig-horseradish peroxidase (HRP) (TAGO) or rabbit anti-mouse Ig-HRP (DAKO) were used in the first step. Swine anti-rabbit Ig-HRP (DAKO) was Materials and methods used, when polyclonal rabbit antibodies were used in the first step. Patients

During the period of 1 April 1983 to 30 November 1985 Results experienced surgical pathologists were unable to classify 29 anaplastic large cell tumours, obtained from 15 females and The staining combinations which can be met in practice 14 males, between 47 and 82 years of age. The tissue with our antibody panel are given in Fig. 1 and indicated specimens were either derived from patients of our hospital, A, B, C, D and E. Their relevance to reach a definite or were sent to our laboratory as a diagnostic problem. diagnosis is also given. Pleiomorphic cells were found to be the main cell type in In table 2, the biopsy location, the differential diagnosis 21 tumours, round cells in 3 tumours, and spindle cells in based on conventional histology, the immunohistochemical 5 tumours. The differential diagnoses, based on reaction patterns obtained with the panel, the diagnosis conventional histology, comprised 21 cases of carcinoma based on it, the verification by additional data as well as or large cell lymphoma, 2 cases of carcinoma or sarcoma, clinical follow-up are shown. and 6 cases of carcinoma, sarcoma or melanoma (Table 1). In 21 cases the differential diagnosis made on routinely processed paraffin sections concerned carcinoma versus Tissue processing lymphoma (no. 1-21 of Table 1). In 17 ourt of these 21 cases combination D was found (anti-keratin-; anti-virnentin If fresh material was obtained, representative tissue + ; 115D8 -; NKI/C-3 -; anti-CLA +), indicating that the samples were (a) snap frozen in liquid nitrogen (b) fixed tumours was lymphomatous by nature (see also Fig. 2). in 2% phosphate buffered glutaraldehyde (pH 7.4) for EM, Subsequently, the diagnosis was confirmed by the presence (c) fixed in Burckhardt's fixative and embedded in of markers for B cell lineage (14 cases) or monocytic lineage methacrylate (Te Velde et al., 1977) for plastic sections and (3 cases). EM was compatible in all cases. Clinical follow- (d) fixed in 4% neutral buffered formalin or a form01 up showed partial or complete tumour regression under sublimate mixture (Bosman et al., 1977) and subsequently poly- (CT) sometimes in combination with embedded in paraffin wax (mpt 52-54oC) for conventional local (RT) or RT alone. Five of these histology. In 4 cases only formalin-fixed and paraffin- patients died within 3-10 months with systemic involvement. embedded material was available. In one of these cases (no. 18) combination E was found For conventional histology sections were stained with (only anti-vimentin +). With this combination a large cell haematoxylin-eosin, Giemsa and PAS whith and without lymphoma or sarcoma is possible. Since on histology diastase digestion. In addition a silver impregnation stain sarcoma was not considered, a large cell lymphoma was and an elastica stain according to van Gieson were made. highly probable. As the tumour cells in addition reacted positive with the markers for histiocytic/monocytic lineage lmmunohistochemistry (FK24, MY7, M02, and lysozyme, AAT and ACT), and markers for B cell (leu 12) and T cell (leu 4, TA1) lineage 4-pm section of the paraffin-embedded tissue were were negative, a true histiocytic lymphoma was indicated. mounted on poly-L-lysine coated glass slides (Huang et al., EM showed cells without desmosomes, abundant cytoplasm 1983), and stained by the indirect immunoperoxidase with varying numbers of lysosomes and phagosomes, and technique. Endogenous peroxidase activity was blocked irregular nuclei. Clinical follow-up showed poor tumour with 0.5% H202 in methanol. For the detection of regresion under CT and local RT. The patient died after and vimentin in formalin-fixed, paraffin- 2 months. In 3 other of these 21 cases (no. 19, 20, 21) embedded tissue, tissue sections were pretreated with 0.1 % combination. A was found (only anti-keratin and 115D8 pronase for 15 min at 37' C (Merkel et al., 1981). Cryostat positive) indicating that the tumour was carcinomatous by sections were also stained by the indirect immunoperoxidase nature (see also Fig. 3). EM and clinical follow-up technique as previously described (van der Valk et al., confirmed this. One of these patients died after 17 months 1983). In each staining series a known positive control (no. 19) with lymph node and cerebral metastases. Fig. 1. Classification of anaplatic large cell tumours by a panel of antibodies HISTOLOGY anaplastic large cell tumour: DD lymphoma, carcinoma, sarcoma, rnelanorna

use panel

Antibody panel:

E (n= l)"

Anti-kerati

Anti-vimentin 1 15D8

NKIIC-3

anti-CLA

definite carcinoma probably carinoma: probably melanoma; definitive lymphoma sarcoma or exlude (histologically exlude histologically lymphoma and clinically): (most cases) or (lymphoblastic, epithelioid sarcoma immunohistochemically centroblasticl * synovial sarcoma (some cases): chemodectoma 1 ' clear cell salcoma neuroblastoma additional immunohistochernistry and electronmicroscopy required for classification

* indicates the number of cases present in this study with this combination 110 lmmunohistochemistry of anaplastic large cell tumours

In 2 cases (no. 22 and 23) on conventional histology these 6 cases (no. 24,25) combination A was found (anti- the differential diagnosis of carcinoma versus sarcoma was keratin and 115DI + ; NKI/ C-3 + or-), indicating a made. In both cases combination B was found (anti-keratin carcinoma. This was confirmed by EM in both cases. and anti-vimentin + , 115D8 + OR-), indicating that the In the other 4 cases combination C was found (anti- tumour might be a carcinoma. Since mesothelioma, vimentin and NKI/C-3 +), indicating a melanoma. Other epithelioid sarcoma and synovial sarcoma, which could also histogenetically related tumours such as chemodectomas, be possible with this combination, could be excluded on neuroblastomas and carcinoids could be excluded with conventional histology and clinical data, the diagnosis conventional histology. Additional immunohistochemistry carcinoma was made. In both cases this diagnosis was showed a positive reaction for S-100 protein in the tumour confirmed in the laryngectomy specimens by conventional cells. EM was confirmative in all 4 cases. Follow-up tissue histology, and (in no. 23) by EM of the original biopsy. specimens showed focally melanin in tumour cells in the In 6 cases the differential diagnosis of carcinoma, Schmorl staining. sarcoma or melanoma was made (no. 24-29). In 2 out of

Fig. 2. Undifferentiated large cell tumour finally clasified as centroblastic lymphoma (case no. 8). a. A high power view of the tumour shows its anaplastic nature and focal tendency to grow is sheets. H.E. X 330. b. Vimentin staining on paraffin section shows a positive reaction in part of the tumour cells (see arrowheads). X 528. c. lmmunostaining with anti-CLA on a cryostat sectioon shows a positive reaction in the tumour cells. X 528

lrnrnunohistochernistry of anaplastic large cell tumours

Fig. 3. Undifferentiated large cell tumour finally classified as on a paraffin section. X 264. le) lmmunostaining with anti- poorly differentiated squamous cell carcinoma (case no. 20). vimentin on a cryostat section shows a positive reaction in (a.b.1 A high power view of the tumour shows its anaplastic fibroblasts and lymphocytes. The tumour cells do not react. nature, and tendency to grow in sheets. H.E. X 264. (c.d.1 X 208. (f.) Electron microscopy shows a detail of two cells lmmunostaining with anti-keratin (C)shows a positive reaction with several desmosomes. No other distinguishing in the tumour cells on a cryostat section and with 115D8 (D) characteristics. X 7,500 Table 1. Clinical and diagnostical data of the patients

AGE ANTI-> ANTI-2 ANTI^ DIAGNOSIS MADE VERIFICATION BY. DEFINITE CLINICAL FOLLOW UP IYEARSl 'EX' 'loPSY KERATIN VlMENTlN CLA 11508~ NKIIC 32 AFTER PANEL ADDITIONAL TECHNIQUES DIAGNOSIS

DD CARCINOMAILYMPHOMA~ 54 M breast lrightl lymphoma EM comp lymph, no ev far ca IH leu 12. Skin: multiple leslons. CT-I - CR. Freeaf disease. Centroblastlc IgM. A HLA-OR + 4 mo. lymphoma

2 79 F tonsil lleftl lymphoma EM: comp lymph, no eu for ca IH: leu12, Localizat~onsin LN, pleura. spleen. CT~I- + 13 centroblastic IgM, A . HLA~DR mal. lymphoma

3 78 F tons11Ilefr! lymphoma EM: comp lymph. no ev for ca IH IgM. A Stainlng procedure, negatwe. RT - CR. Free of centrablast~c dlsease. + 6 mo. lymphama

4 70 M kldnev ileftl lymphoma EMcomp lymph, no eu for ca IH: IgM, A Referred patienl. dlagnasls elsewhere carclnoma centroblastic CT-II - PR. Revision d~agnasisI" lymphoma on lymphama or~ginalbiopsy. Local~zationn retroperltoneal LN 3 mo later CT-IIRT - CR. Free of disease + 22 mo.

5 75 M nasapharynx lyrnphonla EM. comp lymph, no eu lor ca IH: leu12. Staging pocedure: laallzatian ~n cervical LN CT~ centroblast~c IgM. K. HLA~DR IIRT - CR. Free of d~sease. + 4 mo. lymphoma

6 70 M tongue lymphoma EM camp lymph, no ev for ca IH: leu 12. Staging procedure: negative. RT - CR Free of centrablastlc IgG, K. HLA~DR disease. + 5 mo. lymphoma

7 74 F publlc 0s lymphoma EM comp lymph, no ev for ca IH: leu 12. Several pelvic lacalizations CT-IIRT - CR Free centroblast~c IgM. K. HLA~DR of disease. + 16 mo. lymphoma

8 82 F nasal cavlty irtghtl" lymphama EM: camp lymph, no ev lor ca IH: leul2. Stagmgprocedure. negative. RT - CR. After 16 centrabastmc IgM, K. HLA~DR mo localizat~anin retroperitaneal. ingwnal LN. CT-I lymphama - + 13 mo.!

9 70 M nasopharynx lymphoma EM: comp lymph, no ev for ca IH leu12. Staging procedure: negative. CT-IIRT - CR. Free cenlroblaslic tgM. A . HLA-DR of dlsease, + 6 mo. lymphoma

!0 74 F maxilla lrlghtl lymphoma EM comp lymph, no eu for ca IH: leu12, Staging procedure: localization in cervical LN. CT- centroblastic IgG. K, HLA~DR liRT - CR. Free of dosease + 14 mo. lymphama

l 74 F nasopharynx lymphoma EM: camp lymph, no eu for ca IH: leu12, Staglng procedure. lacal~~atianeIn cerv8cal. IgG. A. HLA DR mguinal LN. CT IIRT - CR. Free of disease. + cenlroblasric 14 mo. lymphama

!2 82 F buttock lrtghti Iymphoma EM: comp lymph, no ev for ca IH. leu12, Other local~zat~onin pleural cavity. CT I - PR - cenrroblastic IgM. K. HLA CIR cerebral matastas~s - + 110 m01 lymphoma

'3 M M nasapharynx lvmphoma EM comp lymph, no ev far ca IH leu12. Other lacal!zat~anIrl cerv~calLN CT~I- CR. Free centroblastfc IgM, K. HLA DR of dlsease. + 8 mo. lymphom a

!4 68 F cerv~callymph node iymphoma EM: comp lymph. no ev for ca \H leu12, Undifferentlated tumour of the tongue 3 years centroblastic IgM. A. HLA DR before Treated elsewhere as carcinoma by lvmphoma resection. Revision diagnosis I" lymphama on arlgmal specimen. Other lacalizat~onm axillar LN. CT-I - + 15 m01

!5 68 F cerv~callymph node - 3 + lymphoma EM: comp THL, noev for ca IHFK 24. my7, Other lacal~zat~on~n medlastinal LN. CT IIRT - true htstiocytic M02. Lys. AAT. ACT. HLA~DR + 15 mol. lymphoma

!6 78 F nasal cavity Irlghtl - 3 + lymphomd EM: comp THL, no eu far ca IH. FK 24, my7. Other lacalizations in cervical LN and EM. CT~I true h~stiacytic M02. AAT. ACT, HLA-DR - CR Free of dlsease. + 20 mo. lymphoma

'7 75 AM lonsil lleftl 3 L lymphomd EM: comp 7HL, no ev for ca IHFK 24, hy7. Other lacal~zat~onin cervtcal LN. CT-I!RT - CR true hvstiacytic M02. Lys, AAT, ACT. HLA~DR Free of disease + 15 mo. lymphoma

!8 58 M maxilla lrlghtl - 3 + lymphama sarcoma EM: comp THL, no ev for ca IH. FK 24, my7, staging procedure: Negalive. CT-IIRT - + I7 true h~stiocytic M02, lys. AAT. ACT. HLA~DR m01 lymphoma

!9 66 M ax~llarlymph node 3 + - carcinoma EM: adena d#fferenr#at#an Adeno carcanoma of the stomach 3 years poorly before. Localmzation in rnedtaslinal, diftrentmated retroperltaneal LN. RT - progressive -- ceiebral aderla carclnoma merastasls - + 117 mol. E E E I E E F P:: 0 0 :: C C m _m - -g Fg Eg 23 3a 2a lmmunohistochemistry of anaplastic large cell tumours

Table 2. Panel of antibodies used for the differentiation of anaplastic large cell tumours

ANTIBODY SPEClFlClTY SOURCE REFERENCE

Anti-keratin (P) cytokeratins, broad spectred. (a) Anti-vimentin (P) vimentin, mol. wt 57 KD. present on mesenchymal cells and sarcoma. (b) 1 15D8 (M) MAM-6: human milk-fat globule membrane antigen, mol. wt 400 KD. Hilkens et al., 1984 Present on nearly all carcinomas NKI/C-3 (M) a melanoma-associated antigen. Recognizes a hterogenous glycoprotein, (c mol. wt 25-1 10 KD. Also present in cells of neuroectodermal origin and Vennegoor et al, 1985 . Anti-CLA (M) clones PD 7/26, 2811, reecognizes a determinant with a mol. wt of 200 (a) KD. Present on nearly all lymphoid tumours (not all lymphoblastic tumours rect).

P: polyclonal; M: monoclonal; CLA, common leucocyte antigen

(a) purchased from DAKOPATTS, Copenhage, Denmark (b) purchased from Eurodiagnostics, Apeldoorn, The Netherlands (C) obtained from the Netherlands Cancer Institute, Amsterdam, The Netherlands

Table 3. Antibodies used to verify the final diagnosis

ANTIBODY SPECIFICITY SOURCE leu4 (M) human T lymphocyte antigen TA1 (M) human T lymphocyte antigen leu12 (M) human B lymphocyte antigen anti-ig, k, A (P) human immunoglobulins, K and A light chains of human immunoglobulins present on human B lymphocytes HLA-DR (M) (OKla) B lymphocytes, activated T lymphocytes, monocytes, histiocytes, Langerhans cells, veiled cells, interdigitating cells. (OKM1) monocytes, histiocytes, granulocytes, subpopulation of T cells and most null lymphocytes normal and leukemic human myeloid cells. The antigen is expressed on monocytes, histiocytes, granulocytes human myeloid antigen. Found on monocytes, histiocytes. blast cells of inyelomonocytic Anti-lysozyme (P) histiocytes, myeloid cells, lacrimal and salivary glands Anti-AAT (P) histiocytes, yolk sac tumours, foetal liver, hepato cellular carcinoma Anti-ACT (P) histiocytes Anti-desmin (P) muscle cells: smooth, skeletal, cardiaque Anti-myoglobin (P) muscle cells: skeletal, cardique Anti-neurofilament (P) mwt 70 KD, neurons Anti-F Vlll Rag (P) endothelial cells, megakaryocytes Anti-S 100 protein (P) supportive cells of nervous tissue, melanomas, Langerhans cells, veiled cells, interdigitating cells

(M): Monoclonal antibody; (P): Polyclonal antibody, mwt: molecular weight a. purchased from Beacton and Dickinson, Monoclonal Centre, U.K. b. obtained from Drs. J.H. Kersey and T.W. Lebien, University of Minesota, U.S.A. C. purchased from Dakopatts, Copenhagen, Denmark d. purchased from Ortho Pharmaceutical Corp. lmmunobilogy div., U.S.A. e. obtained from dr F. Koning, University Hospital Leiden, the Netherlands f. purchased from Coulter Clone, U.K. 9. purchased from Eurodiagnostics B.V., the Netherlands h. purchased from Cappel Lab., U.S.A. k. purchased from Monosan, SANBIO b.V., the Netherlands. 116 lmmunohistochemistry of anaplastic large cell tumours

Discussion , these tumours have to be excluded first (Chase et al., 1984; Corson et al., 1984). In the large majority of In this study we have presented the results obtained the cases this can be done on conventional histology and with a panel of 5 antibodies applied on 29 initially clinical details. This reaction pattern, in which markers unclassifiable anaplastic large cell tumours. Only based on characteristic for mesenchym (vimentin) and this panel we were able to achieve a definite diagnosis in (keratin) are both present, is rather particular, and is 22 cases (76%). When the clinical data and histology were considered to be rare in carcinomas. (Ramaekers et al., included in the assessment, even 28 cases (95%) could be 1983; Rungger-Brandle et al., 1983). However, coexpression classified directly. In only one case additional techniques of vimentin and keratin is described in thyroid carcinomas were needed. (Miettinen et al., 1984; Henzen-Logmans et al., 1987) and Although reports have appeared dealing with the use pulmonary carcinomas (Jasani et al., 1985; Mullink et al., of immunocytochemistry in the differential diagnosis of 1986). If 115D8 reacts positively it further strengthens the some types of these tumours, e.g. carcinoma or melanoma epithelial nature of the tumour cells (Hilkens et al., 1984). (Gatter et al., 1985), carcinoma or lymphoma (Gatter et In reaction-pattern C (keratin-; vimentin + ; 115D8 al., 1984; Lauder et al., 1984; Bosq et al., 1985), our -; NKI/C-3 + ; CLA-) a positive reaction for NKI/C-3 antibody panel has the advantage that the whole range of makes the diagnosis of melanoma especially likely (Duinen anaplastic large cell tumours is covered. This is important et al., 1984; Mackie et al., 1984; Vennegoor et al., 1985). because sometimes more than two possibilities have to be But this reaction pattern can also be found in considered in the differential diagnosis of these tumours. histogenetically related tumours such as carcinoids, Another feature of our study, compared with other reports, chemodectomas, clear cells sarcomas and neuroblastomas is that it is a prospective one and was evaluated by clinical (Vennegoor et al., 1985). They are often easily excluded follow-up. by means of conventional histology alone. If necessary the The panel is composed in such a way that (1) the presence of neurofilament can be helpful to exclude obtained positive and negative reaction patterns reinforce neuroblastoma (Carlei et al., 1984). Clear cell sarcomas of each other, and (2) its antibodies all have a specificiaty for the soft tissues have a similar reaction pattern to certain tumour types (see Table 2) although none of them melanomas, which is in accordance with their kinship. is decisive on its own. The least specific antibody in this (Chung et al., 1983). Anti S-100 protein, and other panel is NKI/C-3 since it is not only present in melanoma melanoma-associated antibodies such as HMW 653.408 cells, but also in histogenetically related cells such as can be used for additional support of this diagnosis (Natali chemodectomas neuroblastomas, carcinoids and et al., 1981; Wen et al., 1983; Ruiter et al., 1986). mesotheliomas Duinen et al., 1984; Mackie et al., 1984; Reaction pattern D (Keratin -; vimentin + 115D8 -; Vennegoor et al., 1985). However, a1 the moment no NKI/C-3 - CLA +) leads to a definite diagnosis of melanoma-associated antibodies with higher specificity, lymphoma. Although CLA is present on leucocytes, in which can be used on paraffin sections, are available. S-100 practice the presence of this antigen in anaplastic large cells protein, another antigen known to be present in melanoma tumours supports the diagnosis lymphoma. Additional cells, and also preserved in paraffin sections is even less markers are necesary to characterize the lymphoid tumours specific (Kahn et al., 1983). Other more specific melanoma- into a B, T or histiocytic cell type (Stein et al., 1984). associated monoclonal antibodies like HMW 653.403 Reaction pattern E (keratin -; vimentin + ; 1158D8-; (Natali, 1981) can only be used on frozen section. NKI/C-3 - CLA -) Makes a sarcoma probable. The various immunohistochemical reaction patterns, Additional markers are needed for subtyping (see Roholl which can be met in practice, are shown in Figure 1, and et al., 1985). Since, in our experience, some lymphoblastic will be discussed now. and centroblastic do not stain for CLA with Reaction pattern A (keratin + ; vimenti-; 115D8 + the monoclonal antibody we used (DAKO-LC) (Meijer et or-; NKI/C-3 + or-; CLA-), leads to a definite of al., 1985) these tumours have to be excluded. They can keratin, and the absence of CLA and vimentin in the easily be excluded histology histologically by their tumour cells. The presence of the HMFGM antigen monotonic cell pattern (most cases) or by the use of MAM-6, as detected by 115D8, gives additional support additional lymphoma markers. to the epithelial nature of the cells. 115D8 is included in In summary, in a prospective study with clinical follow- the panel as it can be used on routinely processed material, up, we could extend previous accounts of the use of without significant loss of antigenicity (Hilkens et al., 1984). immunohistochemistry for the classification of anaplastic Especially if the tissue is not properly fixed, the large cell tumours (Borowitz et al., 1984; Gatter et al., 1984; demonstration of this antigen is often decisive in the final Lauder et al., 1984). Moreover, when used as a small diagnosis. selected panel of antibodies, the whole range of these Reaction pattern B (keratin + ; vimentin + ; 115D8 + tumours can be covered and will directly result in a definite or-; NKI/C-3 + or-; CLA-) makes a carcinoma diagnosis of the vas majority of cases. In only a minority probable since keratin is not present in most sarcomas. or for subtyping (lymphomas!), additional markers have However, as this marker pattern can also be found in to be used. mesotheliomas, synovial sarcomas, and epithelioid lmmunohistochemistry of anaplastic large cell tumours

Ackronowledgernents. We thank Mrs. D.I.M. de Jong and Mrs. T.L. and Meijer C.J.L.M. (1 987). Expression of cytokeratins C.P. van Rijn for typing the manuscript, Mrs. T.M. Tadema, G.J. and vimentin in epithelia1 cells of normal and pathologic Oskam and Mr. J.T. van veldhuisen for their technial help, and thyroid tissue. Virchows Archiv. A, 410, 347-354. Mr. E.M. Boots for his tecnical assistence in electron microscopy. Hilkens J., Buijs F., Hilgers J., Hageman Ph., Calafat A., Sonnnenberg A. and Valk M. van der (1 984). Monoclonal antibodies against human milkfat globule membranes detecting differentiation antigens of the mammary gland and its tumors. Int. J. Cancer, 34, 197-206. References Huang W.M., Gibson S.J., Facer P., Gu J. and Polak J.M. (1 983). Improved section adhesion for Battifora H. (1984). Recent progress in the immunocytochemistry using high moleculas weight immunohistochemistry of solid tumors. Semiars in polymers of L-lysine as a side coating. Histochemistry 77, diagnostic Pathology Vol. 1, 4, 251 -271. 275. Baumal R., Kahn H.J., Bailey D., Philips M. J. and Hanna W. Jasani B., Edwards R.E., Thomas N.D. and Gibbs A.R. (1 985). (1 984).The value of immunohistochemistry in increasing The use of vimentin antibodies in the diagnosis of malignant diagnostic precision of undifferentiated tumours by the mesothelioma. Virchows Archiv. A, 406, 441-448. surgical pathologist. Histochem. J. 16, 1061 -1 087. Kahn J.J., Marks A., Thom H. and Baumal R. (1983). Role Borowitz M.J. and Stein R.B. (1 984). Diagnostic applications of antibody to S100 protein in diagnostic pathology. Am. of monoclonal antibodies to human cancer. Arch. Pathol. J. Clin. Pathol. 79, 341-347. Lab. Med. 108, 101-105. Lauder I., Holland D.Y., Gowland G. and Cunliffe W.J. (1 984). Bosman F., Kreuning J., Wal A.M. van der Kuiper G. and Identification of large cell undifferentiated tumours in lymph Lindeman, J. (1977). The influence of fixation or nodes using leucocyte common and keratin antibodies. immunoperoxidase staining of plasma cells in paraffin Histopathology 1, 259-272. sections of intestinal biopsy specimens. J. Histochem. 53, MaKie R.M., Campbell I. and Turbitt M.L. (1 984).The use of 57-62. NKIlC3 monoclonal antibody in the assessment of benign Bosq J., Gatter K.C., Micheau C. and Mason D.Y. (1 985).Role and malignant melanocytic lesions. Am. J. Clin. Pathol. 37, of immunohistochemistry in diagnosis of nasopharyngeal 367-372. tumours. J. Clin. Pathol. 38, 845-848. Merkel K.H.H. and Zimmer M. (1981). Herpes simplex Carlei F., Polak J.M., Ceccamea A., Maragons P.J., Dahl D., encephalitis. A modified indirect immunoperoxidase Cocchia D., Michetti F., Lezoche E. and Speranza V. technique for rapid diagnosis in paraffin embedded tissue. (1984). Neuronal and glial markers in tumours of Arch. Pathol. Lab. Med. 195,351-352. neuroblatic origin. Virchows Arch. (Pathol. Anat.) A 404, Meijer C.J.L.M., Willemze R., Mullink H., Henzen-Logmans 31 3-324. S.C. and Valk P. van der (1985). Monoclonal antibodies Chase D.R., Enzinger F.M., Weiss S.H. and Langloss J.M. in the histopathological diagnosis of non Hodgkin's (1984). Keratin in epithelioid sarcoma. An lymphomas. Neth. J. Med. 28, 138-1 41. immunohistochemical study. Am. J. Surg. Pathol. 8, Miettinen M., Franssila F., Letho V.P., Paasviuo R. and 435-441. Virtanen 1. (1 984). Expression of intermediate filament Chung E.B. and Enzinger F.M. (1983).Malignant melanoma proteins in thyroid gland and thyroid tumors. Lab. Invest. of soft parts. A reassessment of . Am. 50, 262-270. J. Surg. Pathol. 7, 405-413. Muijen G.N.P. van, Ruiter D.J., Ponec M., Hiskens-van der Corson J.M., Weiss L.M., Banks-Schlegel S.P. and Pinkus G.S. Mey C. and Warnaar S.O. (1984).Monoclonal antibodies (1 984). Keratin proteins and carcinoembryonic antigen in wisth different specificities against cytokeratins. An synovial sarcomas: An immunohistochemical study of 24 immunohistochemical study of normal tissues and tumours. cases. Human Pathol. 15, 61 5-621. Am. J. Pathol. 114-9-1 7. Denk H., Krepler R., Artlieb u., Gabbiani G., Rungger-Brandle Mullink H., Henzen-Logmans S.C., Alons-van Kordelara J.J.M., E., Leoncini P. and Franke W.W. (1983). Proteins of Tadema T.M. and Meijer C.J.L.M. (1 986). Simultaneous intermediate filaments. An immunohistochemical and immunoenzyme staining of vimentin and with biochemical approach to the classification of monoclonal antibodies as an aid in the differential diagnosis: tumors. Am J. Pathol. 110, 193-208. malignant mesothelioma-pulmonary adenocarcinoma. Duinen S.G. van Ruiter D.J., Hageman Ph., vennegoor C., Virchows Archiv B, 52, 55-56. Dickersin G.R., Scheffer E. and Rumke Ph. (1984). Nagle R.B., McDaniel K.M., Clark V.A. and Payne C.M. (1 982). lmmunohistochemical and histochemical tools in the The use of antikertin antibodies in the diagnosis of human diagnosis of amelanotic melanoma. Cancer 53, neoplasms. Am. J. Clin. Pathol. 76, 458-466. 1566-1 573. Natali P.G., lmai K., Wilson B.S., Bigotti A., Cavaliere R., Gabbiani G. and Kocher 0. (1983). Cytocontractile and Pellegrino M.A. and Ferrone S. (1981 1. Structural properties cytoskeletal elements in pathologic Processes. Arch. and tissue distribution of the antigen recognized by the Pathol. Lab. Med. 107, 622-625. monoclonal antibody 653.408 to human melanoma cells. Gatter K.C., Alcock C., Heryet A., Pulford K.A., Heyderman J. Nati. Cancer Inst. 67, 591-595. E., Taylor-Papadimitriou J., Stein H. and Mason D.Y. Osborn M. and Weber K. (1 983).Biology of disease. Tumour (1984).The differential diagnosis of routinely processed diagnosis by intermediate filament typing: a novel tool for anaplastic tumors using monoclonal antibodies. Am. J. surgical pathology. Lab. Invest. 48, 372-394. Clin. Pathol. 82, 33-43. Ramaekers F., Puts J., Moesker O., Kant A., Jap P., and Vooijs Gatter K.C., Ralfkiaer E., Skinner J., Brown D., Heryet A., P. (1983). Demonstration of keratin in human Pulford K.A.F., Hou-Jensen K. and Mason D.Y. (1 985).An adenocarcinomas. Am. J. Pathol. l 1 1, A21 3-223. immunocytochemical study of maligriant melanoma and its Roholl P.J.M., Kleyne J., Elbers H., van der Vegt M.C.D., differential diagnosis from other malignant tumours. J. Clin. Albus-Lutter Ch. and van Unnik J.A.M. (1985). Pathol. 38, 1353-1 357. Characterization of tumour cells in malignat fibrous 118 lmmunohistochemistry of anaplastic large cell tumours

histiocytomas and other soft tissue tumours in comparison Valk P. van der, Jansen J., Daha M.R. and Meijer C.J.L.M. with malignant histiocytes. An immunohistochemical study (1983). Characterization of B-cell Non-Hodgkin's on paraffin sections. J. Pathol. 147, 87-95. hlymphomas. A study using a panel of monoclonal and Ruiter D.J., Broecken E.B., Vennegoor C. nad Ferrone S. heterologous antibodies. Virchows Arch. (Pathol. Anat.) (1986). monoclonal antibodies melanoma-associated A 401, 289-305. antigens Chapter 13: Applications of monoclonal antibodies Vennegoor C., Calafat J. and Hageman Ph. (1985). in tumour pathology. M. Nijhoff publ. to be published 1987. Biochemical characterization and cellular localization of a Rungger-Brandle E. and Gabbiani G. (1983). The role of formalin-resistant melanoma-associated antigen reacting cytoskeletal and cytocontractile elements in pathologic with monoclonal antibody NKIIC-3. Int. J. Cancer 35, processes. Am. J. Pathol. 110, 361-381. 287-295. Stein H., Lennert K., Feller A.C. and Mason D.Y. (1984). Warnke R.A., Gatter K.C. and Falini B. (1983). Diagnosis of lmmunohistological analysis of human lymphomas: human lymphoma with monoclonal antileukocyte correlation of histological and immunological categories. antibodies. N. Engl. J. Med. 309-1275-1281. Adv. Cancer Res. 42, 67-147. Wen D.R., Bhuta S., Herschman R., Gaynor R.B. and Cocharan Te Velde J., Burkhardt R., Kleiverda K., Leenhers-Binnendijk A.J. (1983). S-100 protein: A marker for melanocytic I. and Sonnenfeld W. (1977). Methvlmetacrvlate as and tumours. Ann. N.Y. Acad. Sci, 261-266. embedding medium in histopatholo&. ~istopatholo~~1, 319-330. Accepted November 12, 1986

Top View